Optogenetic control of microbial metabolism offers flexible dynamic control over fermentation processes. The protocol here shows how to set up blue light-regulated fermentations for chemical and protein production at different volumetric scales.
Microbial cell factories offer a sustainable alternative for producing chemicals and recombinant proteins from renewable feedstocks. However, overburdening a microorganism with genetic modifications can reduce host fitness and productivity. This problem can be overcome by using dynamic control: inducible expression of enzymes and pathways, typically using chemical- or nutrient-based additives, to balance cellular growth and production. Optogenetics offers a non-invasive, highly tunable, and reversible method of dynamically regulating gene expression. Here, we describe how to set up light-controlled fermentations of engineered Escherichia coli and Saccharomyces cerevisiae for the production of chemicals or recombinant proteins. We discuss how to apply light at selected times and dosages to decouple microbial growth and production for improved fermentation control and productivity, as well as the key optimization considerations for best results. Additionally, we describe how to implement light controls for lab-scale bioreactor experiments. These protocols facilitate the adoption of optogenetic controls in engineered microorganisms for improved fermentation performance.
Optogenetics, the control of biological processes with light-responsive proteins, offers a new strategy to dynamically control microbial fermentations for chemical and protein production1,2. The burden of engineered metabolic pathways and the toxicity of some intermediates and products often impairs cell growth3. Such stresses can lead to poor biomass accumulation and reduced productivity3. This challenge can be addressed by temporally dividing fermentations into a growth and production phase, which devote metabolic resources to biomass accumulation or product synthesis respectively4. We recently showed that the transition from growth to production in this two-phase fermentation can be induced with changes in illumination conditions5,6,7. The high tunability, reversibility, and orthogonality of light inputs8 offer unique advantages to light-controlled fermentations that are difficult or impossible to replicate with chemical inducers used in dynamical control of conventional two-phase fermentations4,9,10,11.
The blue-light responsive EL222 protein derived from Erythrobacter litoralis has been used to develop several optogenetic circuits for metabolic engineering in Saccharomyces cerevisiae5,7,12,13. EL222 contains a light-oxygen-voltage sensor (LOV) domain that undergoes a conformational shift upon blue light activation (465 nm), which allows it to bind to its cognate DNA sequence (C120)13. Fusing EL222 to the viral VP16 activation domain (VP16-EL222) results in a blue-light responsive transcription factor that can reversibly activate gene expression in S. cerevisiae7 and other organisms14 from the synthetic promoter PC120. Several circuits based on EL222 have been developed and used for chemical production in S. cerevisiae, such as the basic light-activated OptoEXP system7, in which the gene of interest is directly expressed from PC120 (Figure 1A). However, concerns of light penetration at the high cell densities typically encountered in the production phase of fermentations motivated us to develop inverted circuits that are induced in the dark, such as the OptoINVRT and OptoQ-INVRT circuits (Figure 1B)5,7,13. These systems harness the galactose (GAL) or quinic acid (Q) regulons from S. cerevisiae and N. crassa, respectively, controlling their corresponding repressors (GAL80 and QS) with VP16-EL222, to repress gene expression in the light and strongly induce it in the dark. Combining OptoEXP and OptoINVRT circuits results in bidirectional control of gene expression, enabling two-phase fermentations in which the growth phase is induced with blue light, and the production phase with darkness (Figure 2A)5,7.
Using light instead of darkness to induce gene expression during the production phase would greatly expand the capabilities of optogenetic controls but would also require overcoming the light penetration limitations of the high cell densities typically encountered in this phase of fermentation. To this end, we have developed circuits, known as OptoAMP and OptoQ-AMP, that amplify the transcriptional response to blue light stimulation. These circuits use wild-type or hypersensitive mutants of VP16-EL222 to control production of the transcriptional activators Gal4p or QF2 of the GAL or Q regulons, respectively, achieving enhanced sensitivity and stronger gene expression with light12,13 (Figure 1C). OptoAMP circuits can achieve complete and homogeneous light induction in 5 L bioreactors at an optical density (measured at 600 nm; OD600) values of at least 40 with only ~0.35% of illumination (5% light dose on only ~7% of the bulk surface). This demonstrates a higher degree of sensitivity compared to OptoEXP, which requires close to 100% illumination12. The ability to effectively induce gene expression with light at high cell densities opens new opportunities for dynamical control of fermentations. This includes operating fermentations in more than two temporal phases, such as three-phase fermentations, in which growth, induction, and production phases are established with unique light schedules to optimize chemical production (Figure 2B)12.
Figure 1: Optogenetic circuits for dynamic control of S. cerevisiae. The OptoEXP, OptoINVRT, and OptoAMP circuits are based on the light-sensitive VP16-EL222 system. (A) In the OptoEXP circuit, exposure to blue light causes a conformational change and dimerization of VP16-EL222, which exposes a DNA-binding domain and allows for transcription from PC120. The figure has been modified from Zhao et al.7. (B) OptoINVRT circuits harnesses the GAL (shown) or Q regulons to induce expression in the dark. In GAL-based circuits, VP16-EL222 and GAL4 are constitutively expressed, while PC120 drives expression of the GAL80 repressor (in Q-based circuits, GAL4 and GAL80 are replaced by QF2 and QS, respectively, and a synthetic QUAS-containing promoter is used instead of a GAL promoter). In light, Gal80p prevents activation of the gene of interest from PGAL1. In the dark, GAL80 is not expressed and rapidly degraded by fusing it to a constitutive degron domain (small brown domain), which allows for activation of PGAL1 by Gal4p. The figure has been modified from Zhao et al.5. (C) OptoAMP circuits also use VP16-EL222 to control the GAL (shown) or Q regulons. In these circuits, the GAL80 repressor (or QS) is constitutively expressed and fused to a photo-sensitive degron (small blue domain) ensuring tight repression in the dark. PC120 and a hypersensitive VP16-EL222 mutant control expression of GAL4 (or QF2) with light, which strongly activates PGAL1 (or a QUAS-containing promoter) in the light. GAL-derived circuits can use engineered forms of PGAL1, such as PGAL1-M or PGAL1-S, which have increased activity, as well as wild-type promoters controlled by the GAL regulon (PGAL1, PGAL10, PGAL2, PGAL7). The figure has been modified from Zhao et al.12. Please click here to view a larger version of this figure.
Figure 2: Two- and three-phase fermentations through time. (A) Two-phase fermentations operated with inverted circuits consist of a light-driven growth phase and a dark production phase. In the growth phase, biomass accumulates as the production pathway stays repressed. Upon reaching the desired OD600, cells are shifted to the dark to metabolically adjust before being resuspended in fresh media for the production phase. (B) In a three-phase process, the growth, incubation, and production phases are defined by unique light schedules, which may consist of a dark growth period, pulsed incubation, and fully illuminated production phase. Figure created with Biorender. Please click here to view a larger version of this figure.
Optogenetic circuits have also been developed for dynamical control of chemical and protein production in E. coli. OptoLAC circuits control the bacterial LacI repressor using the light-responsive pDawn circuit, which is based on the YF1/FixJ two-component system6 (Figure 3). Similar to OptoINVRT5, OptoLAC circuits are designed to repress gene expression in the light and induce it in the dark. Expression levels using OptoLAC circuits can match or exceed those achieved with standard isopropyl β-d-1-thiogalactopyranoside (IPTG) induction, thus maintaining the strength of chemical induction while offering enhanced tunability and reversibility6. Therefore, OptoLAC circuits enable effective optogenetic control for metabolic engineering in E. coli.
Figure 3: OptoLAC circuits for dynamic control of E. coli. The OptoLAC circuits adapt the pDawn system and lac operon to achieve activation in the dark and repression in the light. In the dark, YF1 phosphorylates FixJ, which then activates the PFixK2 promoter to express the cI repressor. The cI repressor prevents expression of the lacI repressor from the PR promoter, which permits transcription of the gene of interest from a lacO-containing promoter. Conversely, blue light reduces YF1 net kinase activity, reversing FixJ phosphorylation and thus cI expression, which derepresses expression of lacI and prevents expression from the lacO-containing promoter. The figure has been modified from Lalwani et al.6. Please click here to view a larger version of this figure.
We describe here the basic protocols for light-controlled fermentations of S. cerevisiae and E. coli for chemical or protein production. For both yeast and bacteria, we first focus on fermentations with a light-driven growth phase and a darkness-induced production phase enabled by OptoINVRT and OptoLAC circuits. Subsequently, we describe a protocol for a three-phase (growth, induction, production) light-controlled fermentation enabled by OptoAMP circuits. Furthermore, we describe how to scale up optogenetically controlled fermentations from microplates to lab-scale bioreactors. With this protocol, we aim to provide a complete and easily reproducible guide for performing light-controlled fermentations for chemical or protein production.
1. Light-controlled chemical production using the S. cerevisiae OptoINVRT7 circuit
2. Light-controlled protein production using the E. coli OptoLAC system
3. Three-phase fermentation using the S. cerevisiae OptoAMP system
4. Chemical (mevalonate) production from E. coli in a light-controlled bioreactor
Optogenetic regulation of microbial metabolism has been successfully implemented to produce a variety of products, including biofuels, bulk chemicals, proteins, and natural products5,6,7,12,13. Most of these processes are designed for cell growth to occur in the light (when low cell density poses minimal challenges with light penetration), and for production to be induced by darkness once the cells are grown. Various chemicals that have been produced from yeast using this approach, including lactic acid, a valuable polymer precursor and food additive, as well as isobutanol, a next-generation biofuel. For both chemicals, a common challenge stems from the strong drive of S. cerevisiae to metabolize glucose toward ethanol production rather than the product of interest and the inability to delete the ethanol fermentation pathway without causing a severe growth defect7. A combination of light-activated OptoEXP and light-repressed OptoINVRT circuits have been used to selectively activate with light the gene for pyruvate decarboxylase (PDC1) required for ethanol fermentation, and induce the pathway for the desired product in the dark5 (Figure 5A,B). Using this strategy with an optimized cell density of induction (ρs), high titers of both desired chemicals can be achieved (Figure 5C,D), highlighting the value of bidirectional control offered by optogenetics.
While most yeast-based optogenetic processes have centered on a light-driven growth phase and darkness-induced production phase, the recent development of extra sensitive and strong OptoAMP circuits has opened opportunities for light-driven fermentations as well12. These light-driven fermentations are similar to the previously described processes; however, the light schedules are reversed such that production occurs in the light. Furthermore, these circuits allow for the implementation of three-phase processes, which adds more flexibility and control to production compared to the standard two-phase approach. Given the sensitivity and strength of these circuits, these three-phase processes are typically optimized by screening different light schedules in each phase. Optimal light pulses depend on the strain and product of interest. Such circuits have been successfully applied to the production of naringenin, a natural product with therapeutic applications, in addition to lactic acid and isobutanol12 (Figure 6). The increased production of all three chemicals demonstrates the value of optogenetic regulation across a range of pathway complexities as well as the new potential offered by three-phase fermentations.
Beyond these demonstrations in yeast, optogenetics has also been applied to enhance the production of proteins and chemicals in the bacterial workhorse E. coli. Fermentations with this host have followed a light-driven growth and darkness-induced production framework using the OptoLAC suite of circuits6. When used to produce a yellow fluorescent protein (YFP) or transcription factor FdeR, light-controlled production is comparable or superior to the levels achieved with standard IPTG induction, but with easier tunability for intermediate levels of production (Figure 7A). In addition, OptoLAC circuits have been applied to produce mevalonate, an important terpenoid precursor, both at the microplate and bioreactor levels (Figure 7B,C). These selected results give a general overview of the strength, versatility, and tunability of optogenetic regulation for microbial chemical and protein production.
Figure 4: Experimental setup for illuminating plates and bioreactors. (A) 24-well plates can be illuminated while shaking by placing a blue LED light panel approximately 40 cm above the shaker. The light intensity should be measured with a quantum meter to ensure it is between ~80-110 µmol/m2/s. (B) To illuminate a bioreactor, place three light panels in a triangular formation around the bioreactor. As with the 24-well plates, the light intensity should be measured and adjusted to reach ~80-110 µmol/m2/s from all sides. Figure created with Biorender. Please click here to view a larger version of this figure.
Figure 5: Light-repressed production of chemicals from S. cerevisiae. To enhance the production of lactic acid (A) or isobutanol (B), a combination of light- and dark-inducible circuits have been used to selectively activate specific pathways. In both scenarios, the essential ethanol production pathway is induced in the light by controlling PDC1 expression with OptoEXP, while the production pathways are activated in the dark using an OptoINVRT circuit. (C) Production of lactic acid was tested at a range of ρs values with two circuits from the OptoINVRT suite. The OptoINVRT7 version performed best, with an optimal ρs value of 7.0. (D) The OptoINVRT7 version also maximized isobutanol production compared to the other circuit, with an optimal ρs of 8.75.**p < 0.01, ***p < 0.001. Statistics are derived using a two-sided t-test. Data are shown as mean values and error bars represent the standard deviations of four replicates. The figure has been modified from Zhao et al.5. Please click here to view a larger version of this figure.
Figure 6: Light-activated production of chemicals in three-phase fermentations of S. cerevisiae. Fermentations using the OptoAMP circuits can operate with three dynamic phases: growth (i), induction (ii), and production (iii), each defined by different light duty cycles. (A) Biosynthesis of lactic acid is induced in blue light by optogenetic control of LDH expression. (B) Lactic acid production can be optimized by using a pulsed (1 s on/79 s off) light schedule in the growth phase and full illumination in the induction and production phases. (C) Production of isobutanol is induced by optogenetically controlling ILV2 expression. (D) Isobutanol production is optimized using a pulsed growth phase (1 s on/79 s off), fully illuminated induction phase, and pulsed (2 s on/118 s off) production phase. (E) Control of the more complex biosynthetic pathway for naringenin, induced by optogenetically controlling expression of the TAL and PAL genes. (F) Naringenin biosynthesis is best optimized using a pulsed growth (1 s on/79 s off), fully illuminated induction, and dark production phase. *p < 0.05, **p < 0.01, n.s. = no significance. Statistics are derived using a two-sided t-test. Data are shown as mean values and error bars represent the standard deviations of four independent replicates. The figure has been modified from Zhao et al.12. Please click here to view a larger version of this figure.
Figure 7: Optogenetic production of recombinant proteins and chemicals in E. coli. The OptoLAC system has been used to produce both proteins and chemicals at titers that are comparable to, or higher than those reached using chemical induction with IPTG. (A) Optogenetic expression of FdeR is both strong and tunable using a variety of light duty cycles, as resolved and quantified with western blot. (B) A strain engineered for mevalonate production exceeds titers achieved using IPTG induction at the 24-well scale at the optimal ρs value. (C) Optogenetic production of mevalonate in a 2 L bioreactor demonstrates the scalability of production beyond microplates. *p < 0.05. **p < 0.01, ***p < 0.001. Statistics are derived using a two-sided t-test. All data are shown as mean values and error bars represent the standard deviations of biologically independent samples. Data for (A) and (C) represent three replicates, while for (B), the number of replicates from left to right= 4, 4, 6, 3, 4, 4, 4, 4. The figure has been modified from Lalwani et al.6. Please click here to view a larger version of this figure.
Dynamic control has long been applied to improve yields for metabolic engineering and recombinant protein production4. Shifts in enzymatic expression are most typically implemented using chemical inducers such as IPTG21, galactose22, and tetracycline23, but have also been mediated using process conditions such as temperature and pH. Optogenetic control of gene expression eliminates the need for changes to fermentation parameters or media composition, making it an easily applicable alternative to traditional induction strategies. The ease with which light can be turned on or off also offers new capabilities like the rapid and reversible tuning of gene dosage. Furthermore, while these protocols focus on blue light-responsive systems, the existence of optogenetic tools that invert existing light responses24 or respond to other wavelengths of light25,26,27,28,29 offers exciting potential to orthogonally control multiple pathways for an unprecedented level of control. Such advantages make light a versatile new solution for flexible control over microbial fermentations.
Beyond screening for the best colonies, as discussed in the protocols, other parameters such as the cell density at which cultures are switched from growth to production (ρs), lengths of the production and incubation periods, and light duty cycles should also be optimized. The best values of these parameters are product- and strain-dependent and should thus be re-optimized for any new application. For example, pathways involving toxic final products may benefit from higher ρs values, which allow for sufficient accumulation of cells before inducing production6,7, while a weaker but less leaky circuit may favor lower values of ρs to maximize total expression time. Likewise, some pathways and recombinant proteins may benefit from intermediate expression levels, which can be achieved with unique light duties. Furthermore, in the case of fermentations that use OptoAMP circuits, the number of temporal phases can be optimized. While the demonstrated protocol describes a three-phase process, fermentations using these circuits can be controlled with a greater number of phases defined by unique light duty schedules and durations. Thus, a range of these parameters should be tested to optimize performance.
Avoiding sources of light contamination presents an important consideration during experimental setup. For processes that require a delay of light stimulation until later stages (e.g., dark to light fermentations), working in a dark room may be advisable to avoid premature activation of optogenetic systems. In these cases, an inert light source can be applied for visibility during the experimental setup (for example, a far-red light source of ~700 nm when working with blue light-activated systems). An advantage of processes that start with light-induced growth (light to dark) is that initial experimental manipulations can be performed under ambient light with ample visibility.
The ease with which a vast number of light duty schedules can be applied to fermentations presents the opportunity to develop higher-throughput methods to balance biosynthetic pathways and elucidate the optimal conditions that maximize fermentation productivity. Instead of balancing metabolic pathways by testing large numbers of combinatorially assembled constructs having each enzyme expressed by promoters of different strengths, pathways could be balanced by varying gene expression levels using different light duty cycles from a much smaller number of constructs. This obviates the need for more cumbersome experimental setups, such as resuspension in different induction media or serial dilutions for chemical inducers. Optogenetic experiments could potentially even be automated to increase throughput by using in silico controllers to deliver specifically timed or localized light pulses to different sample pools30. However, high-throughput experiments under different light conditions must be sufficiently separated to avoid cross-contamination of light, which can pose spatial constraints. Additionally, the requirement of light stimulation prevents the use of most plate readers and micro bioreactors for continuous measurements. While not yet broadly available commercially, several apparatuses and algorithms have been recently developed for high-throughput and continuous optogenetic experiments, which help to address these spatial constraints31,32,33,34. Thus, despite these limitations, optogenetics offers a huge potential to increase experimental throughput while providing enhanced controllability.
The protocols and video presented here will hopefully lower the barriers for other researchers to adopt optogenetic controls of cellular metabolism and microbial fermentations. Optogenetics is an enabling technology for basic research and biotechnological applications that may benefit from fine-tuned control of gene expressions, such as genetics, molecular and cell biology, metabolism, systems biology, and cybergenetics35,36,37,38. Additionally, optogenetic regulation of gene expression has been demonstrated in other microorganisms such as Bacillus subtilis and Pseudomonas aeruginosa, suggesting that the benefits of light control can be extended to the study and application of diverse species39,40,41. These possibilities highlight the future potential of optogenetics for metabolic engineering, protein production, and other biotechnological applications.
The authors have nothing to disclose.
This research was supported by the U.S. Department of Energy, Office of Science, Office of Biological and Environmental Research Award Number DE-SC0019363, the NSF CAREER Award CBET-1751840, The Pew Charitable Trusts, and the Camille Dreyfus Teacher-Scholar Award.
Light-controlled chemical production using S. cerevisiae | |||
24-well culture plate | USA Scientific | CC7672-7524 | |
Agar powder | Thermo Fisher Scientific | 303991049 | |
Aluminum foil | Reynolds | B004NG90YO | |
BioSpectrometer with μcuvette | Eppendorf | 6135000923 | |
Blue LED panel | HQRP | 884667106091218 | |
EZ-L439 OptoINVRT7 Plasmid | N/A | N/A | See Reference 1 |
Glucose | Thermo Fisher Scientific | 501879892 (G8270-5KG) | |
Microcentrifuge | Thermo Fisher Scientific | 75002403 | |
Microcentrifuge tubes | USA Scientific | 1615-5510 | |
Orbital Shaker | Yamato Scientific America | SOU-300 | |
Petri dish | Celltreat | 229656 | |
PmeI | New England Biolabs | R0560L | |
Quantum meter | Apogee Instruments | MQ-510 | |
Replica-plating device | Thomas Scientific | F37848-0000 | |
Replica-plating pads | Sunrise Science Products | 3005-012 | |
SC-His powder | Sunrise Science Products | 1303-030 | |
SC Complete powder | Sunrise Science Products | 1459-100 | |
Sterile sealing film | Excel Scientific | STR-SEAL-PLT | |
YPD agar plates | VWR | 100217-054 | |
Zeocin | Thermo Fisher Scientific | R25005 | |
Light-controlled protein production using E. coli | |||
6X SDS Sample Buffer | Cepham Life Sciences | 10502 | |
12% Acrylamide protein gels | Thermo Fisher Scientific | NP0341BOX | |
24-well culture plate | USA Scientific | CC7672-7524 | |
Aluminum foil | Reynolds | B004NG90YO | |
BioSpectrometer with μcuvette | Eppendorf | 6135000923 | |
Blue LED panel | HQRP | 884667106091218 | |
Coomassie Brilliant Blue G-250 | Thermo Fisher Scientific | 20279 | |
Electrophoresis cell | Bio-Rad | 1658004 | |
Electrophoresis power supply | Bio-Rad | 1645050 | |
LB broth (Miller) | Fisher Scientific | BP97235 | |
Microcentrifuge | Thermo Fisher Scientific | 75002403 | |
Microcentrifuge tubes | USA Scientific | 1615-5510 | |
NaCl | Thomas Scientific | SX0425-1 | |
OptoLAC plasmids | N/A | N/A | See Reference 2 |
Orbital Shaker | Yamato Scientific America | SOU-300 | |
Petri dish | Celltreat | 229656 | |
Quantum meter | Apogee Instruments | MQ-510 | |
SOC medium | Thermo Fisher Scientific | 15544034 | |
Thermomixer | Eppendorf | 5382000015 | |
Tris base | Fisher Scientific | BP1521 | |
Three-phase fermentation using S. cerevisiae | |||
Same materials as "Light-controlled chemical production using S. cerevisiae" protocol plus the following: | |||
EZ-L580 OptoAMP4 Plasmid | N/A | N/A | See Reference 10 |
Chemical production in a light-controlled bioreactor | |||
Aluminum foil | Reynolds | B004NG90YO | |
Antifoam | Sigma-Aldrich | A8311 | |
Bioreactor with control station | Eppendorf | B120110001 | |
BioSpectrometer with μcuvette | Eppendorf | 6135000923 | |
Bleach | VWR Scientific | 89501-620 (CS) | |
Blue LED panel | HQRP | 884667106091218 | |
BPT tubing | Fisher Scientific | 14-170-15 | |
Glucose | Thermo Fisher Scientific | 501879892 (G8270-5KG) | |
Hydrochloric acid (HCl) | Fisher Scientific | 7647-01-0 | |
M9 Minimal Salts | Thermo Fisher Scientific | A1374401 | |
Microcentrifuge | Thermo Fisher Scientific | 75002403 | |
Microcentrifuge tubes | USA Scientific | 1615-5510 | |
NH4OH Solution | Sigma-Aldrich | I0503-1VL | |
Orbital Shaker | Yamato Scientific America | SOU-300 | |
Quantum meter | Apogee Instruments | MQ-510 | |
SC Complete powder | Sunrise Science Products | 1459-100 |